ICD11 Discussion

[sierra-moxon]

Prefix

icd11.foundation

Name

International Classification of Diseases 11th Revision, Foundation

Homepage

https://icd.who.int/en

Source Code Repository

bioregistry

Description

ICD11 foundation is the ontology underlying the ICD 11 codes.
Separate from a linerization, the single parent hierarchy from ICD 10, where there are chapters (eg Q24), sections (eg Q24.11-23), and actual codes (e.g Q24.23), the icd11.foundation is more like an ontology.

License

No response

Publications

No response

Example Local Unique Identifier

1048673005

Regular Expression Pattern for Local Unique Identifier

^\d+$

URI Format String

http://id.who.int/icd/entity/$1

Wikidata Property

No response

Contributor Name

Sierra Moxon

Contributor GitHub

sierra-moxon

Contributor ORCiD

0000-0002-8719-7760

Contributor Email

[email protected]

Contact Name

No response

Contact ORCiD

No response

Contact GitHub

No response

Contact Email

No response

Additional Comments

No response

[matentzn]

Bringing the ICD 11 discussion back here.

Here was my original comment explaining the matter:
#1331 (comment)

My views were confirmed by the WHO affiliate here:
#1331 (comment)

Here is an example of an ICD 11 linearisation (the thing you will see in EHRs):

https://icd.who.int/dev11/l-m/en#/http%3a%2f%2fid.who.int%2ficd%2fentity%2f2134365487

The code is AB12, so we would typically refer to it as icd11:AB12 which is analogous what we were doing with icd10. This is made slightly more complex by the fact that there are multiple linearisations (MMS, dermatology specific) so I would have, in the future, suggestion to prefer icd11.mms or icd11.derma as synonyms, hoping that icd11.mms:ABC == icd11.derma:ABC (the same code ABC refers to the same disease, but they live in different terminologies). That was not always true for icd10, where icd10who and icd10cm did have diverging identifiers occasionally. All of this is not super important for this issue other than to say, careful of using some general prefix like icd11.

For this case here, icd11.foundation lives in an entirely different identifier space.

Here is a link to an old version of the foundation: https://github.com/monarch-initiative/icd11/releases/latest/download/icd11foundation.owl.

Lets look a gouty synovitis:

https://icd.who.int/dev11/l-m/en#/http://id.who.int/icd/entity/1957497923

Its true that there are endpoints that can resolve the one with the other, but that is it.

• the foundation has often different classification than any given linearisation
• the foundation has [0-9]+ as identifier regex, while the linearisation has something along the lines of [A-Z]+[0-9]+
• there are codes that do not correspond to a foundation concept (post coordinated)

I hope this makes the case sufficiently to keep these two things separate..

[cthoyt]

@matentzn this is sounding like an HGNC identifier vs. HGNC symbol situation

A few follow-up questions:

1. Can you please define the word "linearisation" at least once in every thread where you mention it? For the uninitiated, this might cause confusion
2. The page you showed is the MMS interface, which is not the foundation interface, so I'm wondering if there's a better name than "foundation" for these codes? Perhaps icd11.code would fit better. Am I misunderstanding, that there are many different ICD code spaces but there are tiny deviations?
3. How was the OWL file hosted in the Monarch repo that you linked generated? The repository itself doesn't appear to mention what this file is or where it comes from
4. Is there a way to download the list of all the ICD codes? I know that ICD has an API that lets you get the entities, which I dumped to OBO in https://github.com/biopragmatics/obo-db-ingest/tree/main/export/icd11. update: I figured this out in part, see Import ICD11 MMS codes pyobo#285

[matentzn]

Can you please define the word "linearisation" at least once in every thread where you mention it? For the uninitiated, this might cause confusion

A linearisation is a conceptual subset derived from the the ICD-11 foundation component organized into a linear, hierarchical structure (in particular, single inheritance) tailored for specific uses

The page you showed is the MMS interface, which is not the foundation interface, so I'm wondering if there's a better name than "foundation" for these codes? Perhaps icd11.code would fit better. Am I misunderstanding, that there are many different ICD code spaces but there are tiny deviations?

The codes are not "foundation", they are linearisation. I think icd11.code instead of icd11.mms might be fine. So I would be inclined to agree:

• icd11.foundation for foundation iris like http://id.who.int/icd/entity/1957497923
• icd11.code for any code like icd11.code:AB85.
• icd11 being a synonym for icd11.code.

How was the OWL file hosted in the Monarch repo that you linked generated? The repository itself doesn't appear to mention what this file is or where it comes from.

https://icd11files.blob.core.windows.net/foundationowl/whofic-2024-01-21.owl.gz (I don't know whether this is a permanent fixture or if this was done in an ad-hoc manner for us back then to align Mondo with ICD11).

[cthoyt]

I think we agree that icd.code should be its own independent prefix, after considering biopragmatics/pyobo#28.

We can have a follow-up discussion about what icd should mean after, since this requires making an assessment of what resources would be impacted to get the meaning of icd changed

[matentzn]

So you are saying that whatever icd11 is right now, it means icd 11 foundation.

Then I guess let's at the very least add icd11.foundation as a synonym to icd11 for now; I will make it the preferred synonym in the obo context and then you can see how you want to move forward?

[cthoyt]

@matentzn yes, and if you could update the icd11 prefix to explain what "foundation" means and why this is here, it will be helpful for all future ICD11 users

[bgyori]

I'm glad we're all on the same page now! Is it worth redefining icd11 to represent ICD11 codes or remain as a representation of ICD11 foundation IDs? My take is probably no, due to downstream effects but it's a possibility.

[matentzn]

I send this to the ICD 11 WHO team, lets see what they say: